BACKGROUND The aim of the present study was to evaluate whether the level of expression of tissue or plasma miR-106b can be used to predict clinical outcomes in breast cancer patients. METHODS Both tissue and plasma samples were collected and analyzed from 173 patients with primary breast cancer and a set of 50 women with fibroadenoma. The relative expression levels of miR-106b were determined using real-time quantitative reverse transcription polymerase chain reaction and in situ hybridization. RESULTS The levels of miR-106b were upregulated in both tissue and plasma samples from breast cancer patients. The expression levels showed a linear correlation (rs = 0.748, P < 0.001) and were significantly correlated with tumor size, Ki67 expression, and lymph node metastasis (all P < 0.05). Patients with high miR-106b expression levels tended to have shorter disease-free survival times and overall survival times (P < 0.001). In a Cox regression model, high-level tissue and plasma miR-106b expression were unfavorable prognostic factors, and receiver-operating characteristic analysis revealed that the tissue and plasma miR-106b levels provided considerable diagnostic accuracy, yielding an area under the ROC curve of 0.785 and 0.856, respectively. CONCLUSIONS MiR-106b was found to be associated with a high risk of recurrence of breast cancer, and miR-106b is a putative plasma marker for risk assessment in patients with breast cancer.